The UK government has announced a blueprint for artificial intelligence regulation that would allow new AI products to be tested under relaxed rules, in a bid to drive growth and innovation in sectors such as healthcare and housebuilding.
Under the plans, unveiled by the UK's technology secretary Liz Kendall at the Times Tech Summit in London on 21 October, a proposed AI Growth Lab, or sandbox, would enable companies and innovators to test AI tools "in real-world conditions, with some rules and regulations temporarily relaxed under strict supervision".
These new testing environments would initially be set up for key economic sectors including healthcare, transport, and in the use of robotics in advanced manufacturing to "accelerate the responsible development and deployment of AI products", the government said.
It gave the example of a testing ground focused on building AI tools that could "support health workers to deliver better patient care on an accelerated timeline" and reduce NHS waiting lists, and another that could speed up planning approvals and help the government deliver its plan to build 1.5 million new homes by the end of the current Parliament.
At present, a typical housing development application consists of around 4,000 pages of documentation and can take up to 18 months to approve after submission.
Kendall said that to deliver on its plan for national renewal, the government would "need to overhaul the old approaches which have stifled enterprise and held back our innovators".
"This isn't about cutting corners - it's about fast-tracking responsible innovations that will improve lives and deliver real benefits."
GGF report: A fresh mandate for digital leadership in the NHS
Based on interviews with chief digital and information officers in NHS trusts in England, the research highlights the systemic enablers needed to deliver NHS digital transformation - from prioritising digital as a strategic profession to resetting the role of the digital centre.
On AI, it finds that trusts are using or exploring AI across back-office or clinical domains, with one interviewee saying: "We are trying to get our arms around AI before it gets its arms around us."
The report urges trusts to progress with AI in a structured way and avoid introducing further digital fragmentation - a key challenge identified by trusts. Download the report
The overall aim of the AI Growth Lab is to foster economic growth for the UK through responsible AI innovation by making what the government called "targeted regulatory modifications under robust safeguards and with careful monitoring".
It said that regulations would not be "switched on or switched off at will", but would "see strict, time-limited restrictions being put in place to set out which specific regulatory hurdles could be avoided or modified under close supervision".
The AI Growth Lab would be overseen by tech and regulatory experts and "backed up by a strict licensing scheme with strong safeguards", it said, adding that: "Any breaches of individual agreements, or the emergence of unacceptable risks would stop testing in its tracks and open users who have breached their terms up to potential fines."
Read more: Notes from Liverpool: What have we learned from the Labour Party Conference about the UK government's evolving approach to data and AI?
The government said close working between businesses and regulators had already delivered transformations for the public. For example, a sandbox led by the Information Commissioner's Office has supported age verification company Yoti to fine tune their age estimation technology to help keep young people safe online, while another trial has helped FlyingBinary to develop online services which support mental health patients.
It has also reserved £1m (US$1.3m) to support the Medicines and Healthcare products Regulatory Agency to pilot AI-assisted tools. If successful, the tools could support "scientific expertise, speed up drug discovery and clinical trial assessments, and licensing to improve efficiency and consistency".
The government stressed that all decisions made with such tools would remain "in human hands".
The government noted in its announcement of the blueprint for AI regulation that a number of EU countries, the US, Japan, Estonia and Singapore had already announced or implemented sandbox testing grounds for AI.
It said the global appetite to lead in AI meant the UK would need to work to remain "at the vanguard of international best practice in regulatory innovation".
On the same day as the blueprint announcement, the government issued a call for evidence on the AI Growth Lab. The deadline for responses is early January 2026.
The OECD estimates that AI could improve UK productivity by as much as 1.3 percentage points every year - worth the equivalent of £140bn (US$187bn). Currently, only 21% of UK firms are using the technology.
Read more: Australia urged to set 'rules of the game' to harness the productivity potential of AI